Image classification using Deep Ensemble Learning and Transfer Learning methods is performed on a small, labeled dataset of multimodal nonlinear optical microscopy images coming from Stimulated Raman Scattering, Two Photon Excited Fluorescence and Optical Transmission, to differentiate proliferating cancer cells from senescent ones, a peculiar phenotype following an anti-cancer treatment responsible for tumour relapse. The superior performances of the Deep Ensemble Transfer Learning approach are compared with other less complex neural network architectures. Ultimately, the predictions of the neural network are evaluated using the Grad-CAM visualization approach, which allows highlighting the most important features in the input images responsible for the labels assigned by the network.
Recent oncology research highlights that senescence, once deemed beneficial in cancer treatments, can contribute to cancer relapse. Detecting therapy-induced senescent cells is challenging due to their complexity and lack of specific markers. Nonlinear optical (NLO) microscopy provides a fast, non-invasive, label-free detection solution. To distinguish between senescent and proliferating cells, here we present the development of a deep learning architecture based on multimodal NLO microscopy images coming from Stimulated Raman Scattering, Two Photon Excited Fluorescence and Optical Transmission. Despite limited labeled data, Transfer Learning, Data Augmentation, and Ensemble Learning techniques allowed us to achieve an accuracy over 90%. Ultimately, the predictions of the neural network are evaluated using the Grad-CAM visualization approach, which allows highlighting the most important features in the input images responsible for the labels assigned by the network. This work reveals the effectiveness of deep learning in senescence classification, potentially advancing treatment strategies.
Recently, anticancer treatments were discovered to induce cell senescence other than death, a critical phenotype driving tumor recurrence. This calls for the development of safe, precise, and rapid tools to reveal critical therapy-induced senescence (TIS). Here, we present label-free multimodal nonlinear optical (NLO) microscopy as a powerful technique to spot early TIS. We home-built a microscope including different NLO modalities: Stimulated Raman Scattering (SRS), forward and epi-detected Coherent Anti-Stokes Raman Scattering (CARS and E-CARS), and Two-Photon Excited Fluorescence (TPEF). The infrared laser source outputs synchronized narrowband 780 nm pump pulses and 950-1050 nm tunable Stokes pulses, so to match the CH-stretching region of the Raman spectrum. Thanks to the co-registration of these NLO signals from label-free TIS cells and controls, we unveiled quantitative all-optical traits of early-stage TIS, monitored over 72 hours of treatment. TPEF from metabolic coenzymes combined with E-CARS from cardiolipin and cytochrome C indicated an shrinking of mitochondrial networks. CARS and SRS revealed lipid vesicles accumulation in cytoplasms. Nuclei enlarged irregularly, visualized via subtraction of SRS signals of proteins and lipids, and CARS from deoxyribose. We believe our results will strongly influence anticancer pre-clinical studies and translated clinical applications, constituting a quick, non-invasive, and accurate aid to expose TIS manifestation in tumors.
Recent studies have shown that common anticancer treatments can induce cell senescence rather than death, a critical phenotype governing tumor recurrence. This calls for the urgent development of safe, precise, and quick tools to unveil critical Therapy-Induced Senescence (TIS). Merging different coherent Raman and multiphoton techniques, we present label-free multimodal nonlinear optical (NLO) microscopy as a powerful tool to spot early TIS. We home-built a microscope including different NLO modalities: Stimulated Raman Scattering (SRS), forward and epi-detected Coherent Anti-Stokes Raman Scattering (CARS and E-CARS), and Two-Photon Excited Fluorescence (TPEF). The infrared laser source outputs synchronized narrowband 780 nm pump pulses and 950-1050 nm tunable Stokes pulses, so to match the CH-stretching region of the Raman spectrum. Thanks to the co-registration of these diverse techniques applied on label-free TIS cells and controls, we exposed quantitative hallmarks of early TIS, confirmed by comparing different optical signals monitored over 72 hours of treatment. TPEF from metabolic coenzymes combined with E-CARS from cardiolipin and cytochrome C indicated an early shrinking of mitochondria. CARS and SRS revealed lipid vesicles overproduction and accumulation. Nuclei enlarged irregularly, visualized via subtraction of SRS signals of proteins and lipids, and CARS from deoxyribose. We consider our results will strongly influence anticancer pre-clinical studies and translated clinical applications, helping to identify quickly, non-invasively, and quantitatively TIS in human tumors.
Cancer research recently revealed that anticancer therapies can cause cell senescence instead of death, a phenotype governing tumor relapse. Developing safe, quick, and precise tools to spot such therapy-induced senescence (TIS) is an urgency.
We present multimodal coherent Raman and multiphoton nonlinear optical microscopy as powerful to unveil TIS, via a home-built microscope including forward-detected Stimulated Raman Scattering, forward and epi-detected Coherent Anti-Stokes Raman Scattering, Two-Photon Excited Fluorescence and Second-Harmonic Generation modalities. We exposed early TIS in human cancer cells, confirmed comparing diverse signals during therapy period.
We consider our findings will strongly influence anticancer practices, helping prevent tumor recurrence.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.